Bill Connelly’s advanced stats profile isn’t convinced Georgia turned a corner Saturday night. Take a look at the team’s percentile performance in its last four games:
- Vanderbilt: 28%
- Florida: 44%
- Kentucky: 46%
- Auburn: 44%
Not exactly the dramatic uptick you were expecting.
According to Bill’s numbers, the defense did play one of its best games of the season (duh), but the offense didn’t fare nearly as well. On the other hand, it was the worst showing of 2016 by Auburn’s offense, although, oddly enough, the Tigers’ defense actually scored out higher than Georgia’s did. I guess I need more time to process the information.
Get The Picture 
13-7
LikeLike
Yeah. Quantifying what we know, whether it’s an 8th grade teacher making a history test or Bill grading football, is no more an exact science than the Law is.
But, they scored an offensive TD. We didn’t. We won.
LikeLike
i hate to be that guy, but the refs costs the offense at least 1 TD.
LikeLike
ding! ding! ding!
LikeLike
..and the D possibly another.
LikeLike
And the two trick plays should have had a few more points on those drives..and we shut it down on the last possession instead of going for three more. This is a situation where stats don’t tell the story. Anytime you hold the ball and eat the other team’s clock time, you’ve done well. The defense and offense actually are on the same team, Billy.
LikeLike
Yep. This was a low scoring game and when the trick plays blew up and the O got no points it suppressed Georgia’s score. But then, this team’s O has gotten fewer points than it should have based on yards gained all season. I would blame Bobo but he’s not here so I blame Chaney.
LikeLike
Looking down a little further, we’re only projected to have a 70% chance of beating Louisiana Lafayette, and 50% of beating Tech.
I hope the players can keep their heads out of the clouds for a couple more weeks because those are not great odds.
LikeLike
For what it’s worth, and I like the work Bill C does in general, their projections and Sagarin’s are not very good indicators of W/Ls or margin of victory. When I say that, nothing really is or we would all win the bets we make on picks in contests we play in. However, I do think ULL is an opponent to be watched, particularly after the showings we have had txhis year. ULL’s QB threw for over 200 last week against GSU, and ran for almost a hundred with a 10 yard average per carry. I know the Eagles aren’t great on defense but we always have issues with mobile QBs.
LikeLike
Everyone does, Mac , since running QBs are considered as a 12th man.
LikeLike
They may not be great on defense, but their rush defense numbers on the season are pretty good. We best not show up with a “line up in the I and run it at them” Nicholls approach and think it’ll eventually work.
LikeLike
True, GSU couldn’t run it at all. They resprted to the pass and had some success. This could be a Nicholls look with UGA being better with the pass.
LikeLike
We beat an S&P+ top 10 team and dropped 3 spots in those rankings. We are behind 3 teams that we beat head to head, including Missouri, which is currently 3-7.
Forgive me if I take Bill C with a hunk of salt.
LikeLike
Georgia moved up five spots, actually, from 70th to 65th.
The Nicholls game will be an anchor on Georgia’s advanced stats for the rest of the season.
LikeLike
My bad. Apologies. I had 62 in my head for some reason and didn’t double check.
LikeLike
No big deal. Reality is they’ve played two real stinkers and failed to hit the ninetieth percentile once this season. That’s pretty much a formula for mediocrity. And I’d have to say Bill’s numbers for the season reflect what we’ve seen on the field.
LikeLike
Yes and no. I refuse to believe we’re worse than the 3-7 Missouri team that we beat.
LikeLike
Mizzou is 62 and UGA is 65…I’m not sure what game you were watching, but converting on 4th down for a game winning TD and a coin flip feel like roughly the same thing.
But I don’t dismiss your statement out of hand…I do think it’s arguable that UGA is getting better as the season goes on…especially on D…while Mizzou is staying the same.
..
LikeLike
Not a prediction here……would be no surprise if we win this saturday by 4 points with all starters playing late in the 4th qtr. and spank the nerds by 2 points. Stats or not, watch the replay again, review the game film or not…..upon first view, it’s who we are in 2016 football. GO DAWGS!
LikeLike
which is yet another reason why Bill’s analysis is flawed. Nichols and games of its ilk simply don’t usually mean much. In our case, it revealed the depth of our OL problems, but we still outgained them by massive amounts and needed an INT and a fumbled punt to nearly give away the game.
Nonetheless, the larger point is that advanced stats should generally exclude the stats from cupcake games because for most of those games, the stats are so skewed as to create a picture of a team that does not exist once it plays teams on its level.
We are NOT a good team on advanced stats because we are not a good team. That shows up in traditional stats as well, which a cursory look at yardage totals and points scored demonstrates that clearly.
None of this is really relevant because the advanced stats do not pick up the fact that our QB is high ceiling player whose inefficiency is exacerbate by his OL and the fact that he is being asked to make decisions in an offense that requires actual knowledge of how to read defenses and run through progressions. The net result is inefficient QB play. Nor do the advanced stats pick up the major improvement in our DL play because our DL has a significant percentage of its snaps taken by true frosh who are still learning what they should be doing and where they should be lined up.
Again, I break down many of these factors every week when comparing teams playing each other, but I think Bill’s S&P is generally very flawed based on its competitive adjustments.
LikeLike
Sooo… Georgia’s defense put in a better performance against a better offense than Auburn’s defense did, but Auburn’s defense ended up with a better “advanced stats” score than Georgia’s?
Methinks them stats ain’t too danged advanced. 😉
LikeLike
Auburn’s defense didn’t lose the game.
LikeLike
I didn’t say they did.
LikeLike
You said Georgia’s defense turned in a better performance than Auburn’s did. From an offensive scoring standpoint, that’s not the case. And last time I checked, that matters the most.
LikeLike
OK, Boss. I can see I’m going to re-exile myself from the comments section again.
LikeLike
Sorry, but I’m not sure why I’ve offended you here. Whatever the reason, I apologize.
LikeLike
No apology is necessary. I made what I thought was an obvious joke, and I thought you were taking it too literally. I overreacted. Mea culpa, sir.
LikeLike
I watched the entire game, UGA’s shut down a much better O than AU’s D did, that is without question. In the 2nd half, AU was completely stoned. Nada…
The game was much closer to being 31-7 if anything. That and F the barn.
LikeLike
But Georgia’s defense won it. When you don’t get a first down for the whole second half, you’ve faced a good defense. Time of possession and total yards say Georgia’ defense was better.
Auburn’s defense didn’t win it.
LikeLike
You make it sound like Bill’s slagging Georgia’s defense. That’s not the case.
As far as “Auburn’s defense didn’t win it.” goes, if White doesn’t throw that pick-six, what’s the final score?
LikeLike
Amateur hour here…but my understanding on part of what may be driving this is the Barn’s ability to keep us out of the endzone on trips inside the 40.
It’s not their D’s fault that they had to face 81 plays to UGA’s 54. And the fact that they only gave up 6 points on 6 trips inside the 40 stands out. We still don’t create explosive plays with the running game…Chubb ran for an 18 yard run in the 1Q and that was the only run all day over 15. Honestly, if you looked at the offense’s performance without looking at the D’s…you would guess there was a strong likelihood that we lost…the only numbers that would give you pause is those total plays and plays per possession.
LikeLike
Yes, snark aside, there’s no getting around the fact that Auburn’s defense surrendered fewer points than Georgia’s did. I didn’t mean to suggest they didn’t play a whale of a game. They did.
However, I will pick nits with your statement about it not being their fault they had to face more plays than Georgia.
LikeLike
Agree with you. If Auburn wanted to face less plays, they could have forced more 3 & outs. You know, like we did 6 times in a row the second half.
LikeLike
Ed Zachery…they had the option to get a first down, they chose not to excercise it. 🙂
LikeLike
That’s right…the amount of time Auburn’s offense possessed the ball had nothing to do with their defense facing so many plays or even being effective in the 2nd half against said number of plays.
eye roll
LikeLike
Our defense was amazing. But so was theirs. They held a team with Nick Chubb, Sony Michel, etc without a touchdown over 40 minutes of possession. That’s staggering.
LikeLike
Refs helped them..so did trick plays. The trick plays weren’t that awful, but if that last one had been intercepted and returned for a TD, we would be in open rebellion.
If one team kicked a 50 yard FG on their first drive to win 3 to 0 but never got another first down against a team that drove the ball to the one on every possession, but fumbled and lost it on the one each time, which defense was better? The one that gave up the fewest points. Means nothing, sometimes…this is one of those times.
LikeLike
I see your point Auburn’s D hold UGA’s O out of the endzone is not the same accomplishment as holding Auburn’s O (which has destroyed teams as of late) from getting a 1st down in the 2nd half.
Averages will hold hold a team down in ranking. However I saw a team that is improving sure Auburn may have been down with injuries but the D is turning around. Auburn’s D was not crippled with key injuries and the UGA ground game looked the best since UNC. The team may not have rounded the corner but is looking to have turned in on the corner. We will only know if the corner was properly apexed once things are rolling in one direction or another.
Further UGA may not have round rounded the corner but as I said appears to moving in the right direction, the win was great to improve the energy around the program players, coaches, staff, fans, and recruits. The energy especially recruiting maybe the most important thing rolling into next season. I have plenty of things to gripe about but this is a big W for this year and the state of the program moving forward.
LikeLike
Anyone one with eyes can tell which defense played better. So if these advanced stats say otherwise they are worthless.
LikeLike
S&P+ is somewhat useful over the course of a season. It’s still too small a sample to be entirely reliable. From game to game, I think it’s fairly useless. I will say that Georgia is close to the same team it was during the bad part of the season, except it’s improved a bit on special teams and defense and done enough to get the wins late that they couldn’t against Tennessee and Vandy. They’ve outgained every opponent but Florida since the Ole Miss debacle, and the defense has been sound other than the two game stretch against Missouri and Ole Miss. If they can shore up the offensive line, this team should be pretty good next year. If the line wasn’t such an anchor, they’d have been pretty good this year. That said, they could still lose to Tech. The offense would have to play better than it has all season to keep it from being a game decided in the 4th quarter.
LikeLike
There are things killing this team, advanced statswise, that people aren’t seeing. The defense is ineffective in the red zone. The offense has been barely north of 30 points only twice this season. Special teams had their first decent game of the season against Auburn — decent, not great. And it wasn’t Georgia’s offense that scored a touchdown Saturday night.
LikeLike
I get that. I just don’t think the defense’s red zone issues are THAT significant, and if you had a line that could block, the offense wouldn’t look so terrible. And against Auburn, I don’t think the 44% number tells us much. I assume it’s caused entirely by the failure to score a TD on offense despite numerous scoring opportunities. But considering moving the ball was going to be a problem with this line against that defensive front, the fact that they were consistently able to drive the ball was a good performance.
LikeLike
So what you’re saying is that if the offense was better, the offense would be better?
LikeLike
Exactly.
LikeLike
My point is that the S&P+ numbers tell you something about the team, but they don’t tell you everything. This team performed better against Auburn than it did against Florida or Kentucky, regardless of what the advanced metrics think of their percentage performance. I would also argue it performed better against Vandy than Florida, but they lost, so it doesn’t make much of a difference.
LikeLike
I hear you…but I think it’s worth pointing out actual performance and expected performance are two different things. The whole point is that if you have 6 chances inside the 40 and score 6 points; you only have 3 plays of over 25 yards and 0 running plays over 20; you score no TDs; you throw a pick in the endzone; your yards per carry is well bellow the national average; you are getting your running back hit behind the line of scrimmage on 24% of your carries …anyway, if we had lost by 6 instead of won by 6, the conversation today would be COMPLETELY different.
LikeLike
As proven by the reaction to the Vandy game. If they pick up that first down and manage to kick a field goal to win 19-17, people would not have been nearly so down on everything leading up to Kentucky. This team has allowed itself no margin for error all season, because the offense isn’t all that good. They’re about 7 or 8 plays on the season from people either screaming for Kirby’s head or calling him a guy who just knows how to win. Keep in mind, if they’d managed to eke out wins over Tennessee and Vandy, they’d be on the verge of going to Atlanta, and they’d still rank around 50th in S&P+. Those two things wouldn’t really jibe.
LikeLike
Preaching to the choir on that one. I’ve said several times…offense is basically the same as last year and results are basically the same, too, despite W-L.
However, what I saw against Auburn…emerging talent on D (Ledbetter is an absolute beast, y’all…takes a good group and pushes it to great…), great game planning and play calling on D…something more than just playing at the level of their talent. That’s different. That’s something fresh that I haven’t seen in a while. It gets me excited.
LikeLike
Georgia is currently 124th nationally in opponent red zone conversion rate. What would you consider significant?
By the way, check out which team is 125th. Last team to get to the opponent’s 20-yard line wins Saturday. 🙂
LikeLike
Two fewer scores and they jump nearly 40 spots, so yeah, I don’t consider it that significant. Red zone opportunities are more significant to me, and Georgia’s okay on that front.
LikeLike
Bill C uses ‘finishing drives’ instead of red zone scoring…larger sample size and he has a bunch of math that shows why it is one of the five factors.
UGA is 90th. I think there is more than just noise there.
LikeLike
Too early for a full season synopsis, but I think it’s fair to say there have been some surprising results thus far. I would have expected the D to be as good or better all year and play like last Saturday. RZ stat is horrible, but we are currently 11th in total D…that’s pretty good but a little off from last year. But we were #3 in RZ D last year, so that’s a real collapse that’s hard just to explain on personnel.
If Lambert is under center, I’m not going to suggest the record is better, but I wonder if his experience in reading D alignments and making check-offs could have improved the OL performance to open up more for Nick and Sony? We’ll never know. He surely can’t throw some of the passes Skinny does, and if our record was going to be the same either way, then good to get Eason a full season of lumps.
The big thing everybody just knew (!) would be better was the little stuff…”attention to detail”. Clock management. Silly mistakes. Penalties that killed drives. Missed tackles and missed assignments. Far fewer ST screw-ups. Sorry, but as much as I never fully believed that was all on coaching anyway, we sure haven’t seen significant improvement either.
Kirby’s got the #3 recruiting class coming in. I think he’s stockpiled some talent on the bench. The #process is probably starting to settle-in within the locker room. No question we’ll be very competitive the next couple of years, but so will the rest of the SEC East.
LikeLike
Agreed Eason is learning and OL looks to be a priority in the next class. Hopefully Smart closes on signing day. I didn’t expect a division win this year, and if the pains we have experienced pay off later, I am all for it.
LikeLike
I’ll also restate something I’ve noted before. Kirby’s emphasis that “practice is more important” is worthy of snark, but I’m coming around to seeing that it does appear to be his core fundamental philosophy. If practice is right, the games take care of themselves. As much as I find a “game day coach” with the ability to make adjustments and outfox the competition (i.e. SOS) I can see where Kirby is going with this, which is Saban’s #process. The Senator’s previous post about S&C and lack of injuries points to this.
I think the SECe was there for the taking, but we’re playing the long game now. Really hope it pays off.
LikeLike
Agreed, and the SECe may have been for the taking but chances of getting the W against Bama are rather low. Further a beat down to Bama could hurt on the recruiting trail and momentum of the program. It wouldn’t surprise me that if UT gets their 2nd chance against Bama, it would have a lasting negative impact sort of like LSU. I don’t like a loss but I have hope in the future, I just want to win out and close on signing day.
Yes, I admire SOS’s ability to adjust, and wanted a good gameday coach. Smart has been frustrating and a joy which is expected. Part of me thinks another OC could have won an extra game or 2. I still believe he ground out Nichols to drill toughness into the team, which in my opinion goes with the emphasis on practice. Granted I don’t think he expected it to be as close as it was. Smart wants to win in the trenches. The D has adjusted and has shut teams down. I think the long game may take a while but if he succeeds will be worth it.
LikeLike
The defense isn’t the same as it was against OlMs…that’s the thing. Averaging doesn’t tell what the present situation is. How many hits on GTP have you averaged since day one?what’s that got to do with what you get today?
LikeLike
Bill’s not averaging ten years of stats. Just one season.
LikeLike
No it isn’t and for multiple reasons, coaching/time in the system and Ledbetter coming to mind.
LikeLike
Not to take much away from the defensive effort vs Auburn, but the Auburn offense helped that effort at least a couple of times, dropped passes.
Senator you seem to be saying this team is a long way from being good, and I agree. The team has done some good things, made some improvements in some areas. But not nearly enough, yet.
It is easy, especially whilst basking in the light of a big win, to look past some problems we still have which may only be correctable with improved personnel.
In a close game our punt game is still inconsistent enough to get us beat if the bad guys’ offense does not screw up, for instance.
LikeLike
Georgia is a work in progress. I don’t think Kirby Smart would disagree with that.
LikeLike
He would be smart to agree. 😀
LikeLike
…of course he will still be Smart if he disagrees. 🙂
LikeLike
That’s because Smart’s smart, otherwise Smart would smart if he disagreed..
LikeLike
Everyone is a Smartazz….. 😛
LikeLike
Agreed–and the glaring weaknesses early on were O-Line and special teams. Special teams have improved (Love the Blankenship kid–fun character). O-line is still hit or miss…but I think they red-shirted 2 bigguns…and have a few more coming. All that and the defense being fun to watch has me excited about what’s to come. That said, it is still a shame that they haven’t done better by Chubb and Michel this season.
LikeLike
Can’t talk about glaring weaknesses early on without talking about the WRs. Godwin and McKenzie are really talented…but limited in what they can do physically. Wims and Ridley coming on has made a HUGE difference. I think it’s worth pointing out that Eason has as many 50 yard passes in the past two games as Mason and Lambert had in their full two seasons starting.
LikeLike
Wow.
LikeLike
Special teams have been one of the biggest drags on the numbers I think, as they’ve spent most of the year in the 120-114 area. A few more decent games and they’ll get better, but probably no better than 70ish.
Defense has been on an upward trajectory for some time, with the painful exception of red zone defense.
LikeLike
I love the advanced stats. This is a prime example of the reason sports teams still need human scouting in addition to the Moneyball stats guys. I saw a team turn in one of its best performances of the year Saturday. My iPhone calculator may not agree, but… hey, Siri, who won the Georgia-Auburn game on Saturday?
LikeLike
Not responding.
LikeLike
Our offense has scored 5 touchdowns in our last 4 games. The offensive coaching is pathetic.
LikeLike
The comment from a fan of one of the teams Chaney used to work for got it right with his “Chaney gets too cute on offense” comment. If you look at it objectively he has been a millstone around the UGA O all season. I hope we get rid of him and hire a new OC in the off season. Otherwise it will be just more of the same next year.
LikeLike
Ask the Falcons if they wish they’d fired Shanahan after the Falcons horrible offensive year last year.
LikeLike
Advanced stats are great if you have the right framework. think Bill’s advanced stats are based on flawed data. Time and again, Bill’s percentile evaluation proves flawed. If I had to guess, it is because his definition of average includes the performances of mid majors etc in his averages. It becomes stark when you read his previews and you see at team Purdue clocking a 95% percentile performance against Toledo and his characterization of it as a “heck of peak!”
I totally respect the work and effort he puts into trying to identify and quantify the impact of all the relevant factors, but the best measure that I have found is looking at ypp and the differential. Obviously, it can understate the difference if a team is running a number of plays, but it more or less tells you how consistent and explosive a team is. Most teams that don’t have high ypp don’t have a lot of explosive plays (see UGA 2015-2016!).
We held a team that came into the game averaging 6.5 ypp to 3.04. That is not just good. It is insanely good. We took a team ranked in the top 15 in yards per play and held them below the worst team in college football on a ypp basis, and we scored on a pick six. At MINIMUM, that must be a near 100% performance on defense.
On the other hand, we averaged 4.5 ypp on offense against a defense that had surrendered about 5 ypp on the season. So a top 20 defense performed like a top 10 defense against us. Hardly surprising and quite frankly, we were not a LOT below average. Say we had a 35th percentile by that scale…if you average the two units, it should come closer to 70%, which
Focus on the fact that for every play in the game, we gained 1.5 more than our opponent. We gave some of it back on the punting differential and missed FG, but the notion that we put up a 44th percentile performance is sheer folly.
LikeLike
ESPN graded the game a 93 close top no1 nationally for the week at 93 ?(iowa upset mich)
LikeLike
http://www.espn.com/college-football/team/fpi?id=61&year=2016
LikeLike
Exactly
LikeLike
LikeLike
take away the gimmie like nichols st and bill c is 5-4 at predicting uga wins or loses.
LikeLike
Does that take into account the point spread?…otherwise, to me, that stat is useless. YMMV.
LikeLike
we rank 88-90 at field position & finishing drives, but 19 & 28 at explosiveness, efficiency and 28 at turnover margin. his system did not predicted wins or losses for us.
bill c’s “system” doesn’t work for UGA, no better than a coin flip..
LikeLike
You probably know this but they give a projected final score, win percentage, and other percentage predictions based on how close they think the score will be. Using those to verify how I feel the point spread winner will turnout, I don’t find them to be very good at all. Certainly would not rank well in the weekly contest here at GTP, but is a data point I look at, along with other prediction sites (primarily for unanimity, or major differences, in thinking.) Short answer, there is no magic pill that I have found.
LikeLike
Not quibbling with the numbers, just noting that the fact that UGA’s percentile performance has been stable the last three games is amusing to me.
LikeLike
After an exhaustive review of various advanced metrics I have been able to determine two things for certain:
13 is more than 7.
I do not care for Auburn.
LikeLike
I like driving my truck.
LikeLike
Auburn sucks.
LikeLike
Wait! Wait! What about all that math?
Anyone want a cookie about now?
LikeLike
On his UGA page he predicted UGA would lose to Auburn (59%) & Mizzou (64%) odds.
Also predicted UGA would beat TN (60%) & Vandy (85%).
Nichols St doesn’t count, that’s a gimme pick.
So he was 5-4 in his picks.
Oops.
LikeLike